Software-Defined

OpenStack Barcelona Voting is Open! Vote for the Brocade sessions you would love to see.

by Jeni Lloyd ‎07-26-2016 11:58 AM - edited ‎07-28-2016 01:07 PM (2,917 Views)

Barcelona Sessions Blog.JPG

 

The OpenStack Summit travels to Barcelona this October. Now is the time to cast your votes for the sessions you would find most valuable. Remember that most sessions are recorded and posted for on demand viewing so even if you won’t be there in person, it is well worth voting for topics that interest you. The sessions listed below include insights into Brocade integrations and advances in OpenStack environments over the past few months. Voting closes August 8 2016.

 

Note that the voting page doesn’t currently enable direct links to sessions. If that function becomes available, I will update this post with direct links. I have include the session numbers as well as titles for now.

 

NFV Bakeoff – Take 2 (14990)

Speakers: Bob Haddleton, Sridhar Ramaswamy, Christopher Young, Beth Cohen

 

Join a high powered panel of NFV experts to see for yourselves what the major networking vendors are doing to create cutting edge services in a virtualized network world. NFV (Network Functions Virtualization) technology has evolved rapidly since 2010 when it was first proposed by the European Telecommunications Standards Institute; in no small part to the OpenStack community’s active support and development efforts.   It is rapidly gaining acceptance in the data center and telecom industries as the hottest technological transformation that allows companies to virtualize networks to cut administrative overhead and increase the flexibility of their network services.  Join some of leading NFV vendors including Brocade, Nokia, Cisco and Juniper as we discuss the reality and challenges of bringing NFV into large scale networks.

 

Hitachi private cloud infrastructure built on OpenStack (15767)

Speakers: Manju Ramanathpura, Henry Chu, Gaurav Singh

 

Unified Compute Platform (UCP) from Hitachi Data Systems is a leading converged infrastructure supporting broad range of enterprise workloads and use cases. With the growing interest, and popular acceptance of OpenStack there has been an increased customer demand to have OpenStack platform validated on UCP, and be made available for Enterprise private cloud, as well as other OpenStack based IaaS cloud deployments. Hitachi Data Systems, Brocade, and Red Hat have developed a validated solution, (UCP) 2000 with Red Hat Enterprise Linux OpenStack Platform 7.0. This solution enables customers to quickly and reliably deploy OpenStack based private and hybrid cloud on converged infrastructure. The solution architecture is built on HDS 2u4n Server for Solution, Brocade 6740 NOS switches, and HDS G400 storage arrays. In addition to that, it includes Red Hat Enterprise Linux 7.3, Red Hat Enterprise Linux OpenStack Platform 7.2, and the Red Hat Enterprise Linux OpenStack platform installer.

 

CTO Perspectives on Service Function Chaining for OpenStack Clouds (15408)

Speakers: Jim Guichard, Joel Halpern, Tom Nadeau, Pere Monclus, Wendy Cartee

 

Service function chaining (SFC) is the underpinning of NFV, enabling operators to compose a dynamic sequence of network services suited for elastic service environments that ultimately determines the quality and efficiency of service offerings. Virtualized network functions is one step in the NFV journey, but they are insufficient without proper SFC. In this session, the panel of CTOS will discuss technologies that are available today and in development for the near future to drive more agile, dynamic SFC with granular service points and policies for OpenStack operators. Some topics to be discussed will include:  

 

  • Role of SFC in NFV  
  • Challenges of deploying SFC
  • Clarifying SFC, Service Insertion, Service Graphs
  • SFC Architecture Framework
  • Standards bodies: ETSI, IETF on SFC
  • SFC working groups and where to find more information

 

Orchestrating VNF Forwarding Graphs and SFC using OpenDaylight, Neutron and Tacker (15335)

Speakers: Sridhar Ramaswamy, Brady Johnson, Louis Fourie

 

Successful deployments of complex Network Services needs dynamic control of traffic flows through VNFs. Even when using SDN, adding new services and changing the related networking in the underlying networks can still be a challenge. Service Function Chaining (SFC) addresses these complexities, making even the most complex scenarios quite simple. SFC implemented in an SDN Controller, like OpenDaylight (ODL), must interact with additional components to automate the entire Service Chaining scenario. A VNF Manager, like Tacker, is necessary to manage the life cycle of VNFs.  Multiple layers need to work in cohesion to make this a reality. This ability to support dynamic VNF Forwarding Graphs using Service Function Chaining differentiates various NFV Orchestration stacks out there in the market.

In this talk, you will hear how three open source projects - Tacker, Neutron and OpenDaylight SDN Controller - coming together to achieve this holy grail of the NFV promise.

 

Tacker 102: Beyond VNF life cycle management (15824)

Speakers: Sripriya Seetharam, Bob Haddleton, Sridhar Ramaswamy

 

Tacker project is an extendable NFV orchestration software that follows a plugin driver model approach for ease of customization and adaptability by network operators. Operators wanting to deploy Tacker in their infrastructure typically write custom drivers based on their VNF requirements and implement end to end service workflows. 

In this session, first, you will learn about Tacker internals on how different components such as VNFM, NFVO, and VIM interact with each other. New features such as alarm based monitoring, manual and auto scaling, events notification and VNF forwarding graphs introduced in the Newton release will be discussed. Then, you will learn how to extend Tacker's capability by writing custom drivers. A step-by-step procedure of writing a new driver for VNF monitoring, configuration and supporting custom VIMs will be presented. Lastly, the audience will learn how to configure the new drivers in Tacker and invoke them through TOSCA VNFD template specification.

 

LBaaS through the Looking Glass (15244)

Speakers: Dhanashri Patil, Ajay Bharadwaj, Matthew Geldert

 

In this session, we zoom in and take a closer look at LBaaS. What were some of the problems we faced in our customer deployments? Learn how we overcame some of the unique challenges in the real world. Take a deep dive into out our enterprise grade LBaaS solution, showcasing high reliability, scale-out and flexible automated licensing. Let us collaborate, think big and leap forward together in the LBaaS journey ahead.

 

Meeting SLA: NFV Notification and Remediation using Tacker

Speakers: Vishwanath Jayaraman, Kanagarajan Manickam

 

The advent of NFV has changed the way TELCOs are beginning to deploy network functions. NFV has brought benefits in terms of deployment agility and cost reduction while also introducing virtualization layers that add complexity and operational challenges. One of the critical requirement for TELCO to maintain SLA is that identifying the abnormal situations of running VNFs and address them to bring it to normal operational state, automatically or manually. This requires monitoring, async event notifications, remediation such as scaling which has been addressed collectively by OpenStack services tacker, ceilometer, heat, etc.

In this session, we will discuss in detail about these features and how OSS can use them automatically to handle the abnormal situations.

 

REST API driven Compute Node configuration for enabling EPA based placement of VNFs by MANO

Speakers: Chegu Vinod, Naziya Khan, Vishwanath Jayaraman

 

VNFD templates allow for specifying requirements for VNF workloads that leverage features of a compute node such as NUMA topology, SR-IOV, PCI-Passthrough, Huge pages and CPU pinning. This allows for Enhanced Platform Awareness (EPA) driven placement of VNF workloads to achieve better performance. However the cloud administrator is required to pre-configure computes to enable for EPA usage. The manual steps involved here are not scalable considering size of typical Telco clouds and can be error prone. This talk introduces a REST-API driven ability to configure/modify compute nodes to enable for EPA usage. OpenStack Tacker (or other ETSI MANO clients) will be able to query which compute nodes are configured for EPA usage and the respective resource allocations. In addition this talk will highlight the need for enhancing/extending OpenStack (e.g. Nova) APIs for retrieving detailed EPA relevant resource usage information from compute nodes.

 

Tacker Newton features hands-on lab

Speakers: Jeff Rametta, Sripriya Seetharam

 

In this session, participants will be provided with an overview of OpenStack Tacker and the basic principles of NFV orchestration. The lab will contain a series of exercises to help users become familiar with Tacker workflow and get hands on experience with some of the advanced service features introduced in the Netwon release.

 

  • Alarm based monitoring
  • Auto-scaling and manual scaling of VNFs
  • VNF Forwarding Graphs using Service Function Chaining (SFC)
  • Eventing mechanism

In this hands on lab the participants will deploy, manage, and monitor complex network services that are self-healing and auto-scalable.

 

NTT West - Beyond Neutron, taking control of a physical multi-vendor network with OpenDaylight

Speakers: Shigeaki Kimura, Jon Castro, Darin Sikanic

 

Neutron provides a rich set of functionality to deliver fully featured networking services between OpenStack tenants. However the roles and responsibilities of Neutron end here, to deliver broader networking capabilities and take programmatic control of the underlying network an SDN controller is of value. OpenDaylight is a popular SDN controller, it provides multiple southbound APIs and a model driven abstraction layer allowing systematic programmability of multi-vendor devices.

Current efforts to integrate OpenDaylight into an OpenStack environment have focused on simply delivering the same functionality as Neutron. In this presentation we will demonstrate how OpenDaylight can be used to configure underlay network elements such as a firewall and expose this ability as a service for OpenStack tenants to take advantage of, essentially allowing physical FWaaS.

 

OpenStack Networking SFC and L2GW service through OpenDaylight

Speakers: Anil Vishnoi (Brocade), Vishal Thaper (Ericsson), Vivekanandan Narasimhan (Ericsson)

 

In this session speakers will mainly talk about the following OpenStack Services and how they are being implemented using the OpenDaylight Controller:

  • OpenStack Service Function Chaining (networking-sfc)
  • OpenStack L2 Gateway Service

Speakers will talk about the architecture and high level design of the solution implemented in OpenDaylight to support the OpenStack SFC API and L2 Gateway APIs. Speakers will go through the details of all the component developed across the OpenStack (networking-odl) and OpenDaylight projects (neutron northbound, netvirt and OpenDaylight SFC project) to realize these services. They will talk about the NFV usecase of OpenStack SFC and dive into details of architecture of how OpenStack Tacker project is consuming OpenStack SFC APIs to realize Tacker VNF-FG API's. Speakers will also go through the details of L2 GW Usecase and discuss the architectural details of the service implementation using OpenStack networking-ODL and OpenDaylight. 

 

Don’t forget! Vote for the Brocade sessions you would love to see before August 8. VOTE HERE